Publication Date


Date of Final Oral Examination (Defense)


Type of Culminating Activity


Degree Title

Master of Science in Civil Engineering


Civil Engineering

Major Advisor

Mandar Khanal, Ph.D.


Kyungduk Ko, Ph.D.


Yang Lu, Ph.D.


Traffic safety may be analyzed with the use of surrogate safety measures, measures of safety that do not incorporate collision data but rather rely on the concept of traffic conflicts. Use of these measures provides several benefits over use of more traditional analysis methods with historical crash data. Surrogate measures eliminate the need to wait for crashes to occur to conduct a safety analysis. The amount of time required for enough crash data to accumulate can be significant, delaying safety analyses. Similarly, these measures allow for safety analysis to be conducted prior to crashes occurring, potentially calling attention to hazardous areas which may be altered to prevent crashes. In addition to these benefits, traffic conflicts occur much more frequently than collisions, generating many more data points which in turn make statistical methods of analysis more effective.

Evaluating surrogate safety measures for a particular transportation network is most effectively done with the use of traffic microsimulation or with connected vehicle data. Traffic microsimulation (such as the use of PTV VISSIM) will generate kinematic data that may then be used for computation of surrogate safety measures. A significant amount of research has been done on this topic, resulting in the establishment of algorithms for calculation of several different surrogate measures and validation of these measures.

Kinematic data from connected vehicles has also been used for the calculation of surrogate safety measures. One data point collected by connected vehicles is harsh braking events which could serve as a surrogate safety measure. Because drivers usually brake more gently if given the opportunity to do so, harsh braking events indicate that a traffic conflict has occurred or is about to occur. Such events take away the driver’s opportunity to brake gently. This research establishes statistical models which relate harsh braking events to crashes on intersections and segments in Salt Lake City, Utah. The findings indicate that harsh braking events have the effect of reducing expected crashes because they represent traffic conflicts which were remedied through the use of harsh braking as an evasive action. The presence of schools and the presence of left turn lanes were also found to be statistically significant crash predictors. In addition to this research work a paper outlining the existing state of safety analysis with surrogate safety measures and evaluating the usefulness and practicality of various existing measures is presented.